Systems and methods for blast electronic activity detection

Abstract

The present disclosure relates to determining detecting blast electronic activities. A method can include identifying a plurality of first electronic activities transmitted by a first electronic account of a data source provider. For each first electronic activity of the plurality of first electronic activities, a plurality of features can be extracted. For at least one first electronic activity of the plurality of first electronic activities, a blast probability score can be generated indicating a likelihood that the at least one first electronic activity is a blast electronic activity. The blast probability score can be generated using a machine learning model trained using features extracted from second electronic activities labeled as blast electronic activities and features extracted from third electronic activities labeled as non-blast electronic activities. An association between the at least one first electronic activity and the blast probability score can be stored in a data structure.

What is claimed is: 
     
       1. A method comprising:
 identifying, by one or more processors, a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generating, by the one or more processors, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generating, by the one or more processors, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; 
 storing, by the one or more processors, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model; 
 identifying, by the one or more processors, a first node profile corresponding to the first electronic account in a node graph and a second node profile linked with the first node profile; 
 identifying, by the one or more processors, a second electronic account corresponding to the second node profile; and 
 transmitting, by the one or more processors, a notification to the second electronic account responsive to the blast probability score being greater than a threshold blast score. 
 
     
     
       2. The method of  claim 1 , further comprising:
 determining, by the one or more processors, a performance score of operation of the first electronic account using the blast probability score associated with the at least one first electronic activity transmitted by the first electronic account. 
 
     
     
       3. The method of  claim 2 , wherein determining the performance score comprises:
 assigning, by the one or more processors, a first weight to each first electronic activity of the plurality of first electronic activities, the first weight determined based on the blast probability score; and 
 determining, by the one or more processors, the performance score using the first weight, the at least one first electronic activity, and at least one fourth electronic activity different from the at least one first electronic activity. 
 
     
     
       4. The method of  claim 1 , wherein the at least one first electronic activity is a particular first electronic activity, and generating the blast probability score comprises:
 assigning, by the one or more processors, at least a first subset of the plurality of first electronic activities to a first cluster using the plurality of first features; 
 determining, by the one or more processors, that the particular first electronic activity is assigned to the first cluster; and 
 generating, by the one or more processors, the blast probability score responsive to determining that the particular first electronic activity is assigned to the first cluster. 
 
     
     
       5. The method of  claim 4 , wherein assigning the first subset to the first cluster comprises determining, by the one or more processors, that each first electronic activity of the first subset has a timestamp within a predetermined period of time, the predetermined period of time having a duration less than or equal to one hour. 
     
     
       6. The method of  claim 4 , wherein generating the blast probability score comprises determining, by the one or more processors, the blast probability score to be relatively greater as a distance between a center of the cluster and a first position of the at least one first electronic activity decreases, the center and the first position defined in a feature space comprising a plurality of dimensions, each dimension corresponding to a respective feature of the plurality of first features. 
     
     
       7. The method of  claim 4 , wherein at least one of (i) the plurality of first features does not include a timestamp, (ii) assigning, by the one or more processors, the first subset of the plurality of first electronic activities comprises assigning, by the one or more processors, a first weight to a timestamp of the plurality of first features that is less than a second weight assigned to another first feature of the plurality of first features; or (iii) assigning the first subset to the first cluster comprises determining, by the one or more processors, that each first electronic activity of the first subset has a timestamp within a predetermined period of time, the predetermined period of time having a duration greater than one day. 
     
     
       8. The method of  claim 1 , wherein generating the plurality of first features comprises generating, by the one or more processors, a first feature data structure comprising a timestamp of the first electronic activity and at least one of a word count of the first electronic activity, a text complexity score of the first electronic activity, a thread length associated with the first electronic activity, a recipient count of the first electronic activity, and a flag indicating whether the first electronic activity is assigned to a cluster of electronic activities. 
     
     
       9. The method of  claim 1 , wherein training the machine learning model comprises:
 applying, by the one or more processors as input to the machine learning model, a training data structure comprising a plurality of training data entries, each training data entry comprising one of (i) a first identifier of a particular second electronic activity of the second electronic activities and the second features of the particular second electronic activity or (ii) a second identifier of a particular third electronic activity of the third electronic activities and the third features of the third electronic activity; 
 causing, by the one or more processors, the machine learning model to generate a candidate output data structure responsive to the input; 
 comparing, by the one or more processors, the candidate output data structure with corresponding first labels indicating that the second electronic activities are blast electronic activities and second labels indicating that the third electronic activities are not blast electronic activities; and 
 modifying the machine learning model responsive to the comparison to satisfy a convergence condition. 
 
     
     
       10. A method comprising:
 identifying, by one or more processors, a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generating, by the one or more processors, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generating, by the one or more processors, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; 
 storing, by the one or more processors, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model; and 
 assigning, by the one or more processors, a first tag to the at least one first electronic activity responsive to the blast probability score associated with the at least one first electronic activity being greater than a threshold blast score; or 
 assigning, by the one or more processors, a second tag different than the first tag to the at least one first electronic activity responsive to the blast probability score associated with the at least one first electronic activity being less than the threshold blast score. 
 
     
     
       11. A method comprising:
 identifying, by one or more processors, a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generating, by the one or more processors, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generating, by the one or more processors, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; 
 storing, by the one or more processors, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model; and 
 assigning, by the one or more processors, a first effort score to the at least one first electronic activity responsive to the blast probability score being greater than a threshold blast score; or 
 assigning, by the one or more processors, a second effort score to the at least one first electronic activity responsive to the blast probability score being less than the threshold blast score, the second effort score greater than the first effort score. 
 
     
     
       12. A method comprising:
 identifying, by one or more processors, a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generating, by the one or more processors, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generating, by the one or more processors, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; 
 storing, by the one or more processors, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model; 
 identifying, by the one or more processors, a candidate record object of a system of record of the data source provider for which a match score with the at least one first electronic activity is greater than a threshold match score; and 
 not matching the at least one first electronic activity with the candidate record object responsive to the blast probability score being greater than a threshold blast score. 
 
     
     
       13. A system, comprising:
 one or more processors configured to execute machine-readable instructions to: 
 identify a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generate, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generate, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; and 
 store, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model 
 assign a first tag to the at least one first electronic activity responsive to the blast probability score associated with the at least one first electronic activity being greater than a threshold blast score; or 
 assign a second tag different than the first tag to the at least one first electronic activity responsive to the blast probability score associated with the at least one first electronic activity being less than the threshold blast score. 
 
     
     
       14. The system of  claim 13 , wherein the one or more processors are configured to:
 determine a performance score of operation of the first electronic account using the blast probability score associated with the at least one first electronic activity transmitted by the first electronic account. 
 
     
     
       15. The system of  claim 13 , wherein the at least one first electronic activity is a particular first electronic activity, and the one or more processors are configured to generate the blast probability score by:
 assigning at least a first subset of the plurality of first electronic activities to a first cluster using the plurality of first features; 
 determining that the particular first electronic activity is assigned to the first cluster; and 
 generating the blast probability score responsive to determining that the particular first electronic activity is assigned to the first cluster. 
 
     
     
       16. The system of  claim 13 , wherein the one or more processors are further configured to:
 identify a first node profile corresponding to the first electronic account in a node graph and a second node profile linked with the first node profile; 
 identify a second electronic account corresponding to the second node profile; and 
 transmit a notification to the second electronic account responsive to the blast probability score being greater than a threshold blast score. 
 
     
     
       17. A non-transitory computer-readable storage medium having instructions embodied thereon, the instructions being executable by one or more processors to perform a method, the method comprising:
 identifying a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generating, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generating, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; 
 storing, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model; and 
 i) a) identifying a first node profile corresponding to the first electronic account in a node graph and a second node profile linked with the first node profile; 
 b) identifying a second electronic account corresponding to the second node profile; and 
 c) transmitting a notification to the second electronic account responsive to the blast probability score being greater than a threshold blast score; or 
 ii) assigning a first tag to the at least one first electronic activity responsive to the blast probability score associated with the at least one first electronic activity being greater than a threshold blast score or assigning a second tag different than the first tag to the at least one first electronic activity responsive to the blast probability score associated with the at least one first electronic activity being less than the threshold blast score; or 
 iii) assigning a first effort score to the at least one first electronic activity responsive to the blast probability score being greater than a threshold blast score or assigning a second effort score to the at least one first electronic activity responsive to the blast probability score being less than the threshold blast score, the second effort score greater than the first effort score; or 
 iv) a) identifying a candidate record object of a system of record of the data source provider for which a match score with the at least one first electronic activity is greater than a threshold match score; and 
 b) not matching the at least one first electronic activity with the candidate record object responsive to the blast probability score being greater than a threshold blast score. 
 
     
     
       18. The non-transitory computer-readable storage medium having instructions embodied thereon of  claim 17 , the method further comprising:
 determining a performance score of operation of the first electronic account using the blast probability score associated with the at least one first electronic activity transmitted by the first electronic account. 
 
     
     
       19. A system, comprising:
 one or more processors configured to execute machine-readable instructions to:
 identify a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generate, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generate, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; 
 store, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model; and 
 assign a first tag to the at least one first electronic activity responsive to the blast probability score associated with the at least one first electronic activity being greater than a threshold blast score; or 
 assign a second tag different than the first tag to the at least one first electronic activity responsive to the blast probability score associated with the at least one first electronic activity being less than the threshold blast score. 
 
 
     
     
       20. A system, comprising:
 one or more processors configured to execute machine-readable instructions to:
 identify a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generate, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generate, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; and 
 store, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model; and 
 assign a first effort score to the at least one first electronic activity responsive to the blast probability score being greater than a threshold blast score; or 
 assign a second effort score to the at least one first electronic activity responsive to the blast probability score being less than the threshold blast score, the second effort score greater than the first effort score. 
 
 
     
     
       21. A system, comprising:
 one or more processors configured to execute machine-readable instructions to:
 identify a plurality of first electronic activities transmitted by a first electronic account of a data source provider; 
 generate, for each first electronic activity of the plurality of first electronic activities, a plurality of first features extracted from the respective first electronic activity; 
 generate, for at least one first electronic activity of the plurality of first electronic activities, using the plurality of first features extracted from the plurality of first electronic activities, a blast probability score indicating a likelihood that the at least one first electronic activity is a blast electronic activity, the blast probability score generated using a machine learning model trained using second features extracted from second electronic activities labeled as blast electronic activities and third features extracted from third electronic activities labeled as non-blast electronic activities; 
 store, in one or more data structures, an association between the at least one first electronic activity and the respective blast probability score determined using the machine learning model; 
 identify a candidate record object of a system of record of the data source provider for which a match score with the at least one first electronic activity is greater than a threshold match score; and 
 not match the at least one first electronic activity with the candidate record object responsive to the blast probability score being greater than a threshold blast score.